Targeted gene disruption has wide applicability for research, therapeutic, agricultural, and industrial uses. One strategy for producing targeted gene disruption is through the generation of double-strand DNA breaks caused by site-specific endonucleases. Endonucleases are most often used for targeted gene disruption in organisms that have traditionally been refractive to more conventional gene targeting methods, such as algae, plants, and large animal models, including humans. For example, there are currently human clinical trials underway involving zinc finger nucleases for the treatment and prevention of HIV infection. Additionally, endonuclease engineering is currently being used in attempts to disrupt genes that produce undesirable phenotypes in crops.
The homing endonucleases, also known as meganucleases, are sequence specific endonucleases that generate double strand breaks in genomic DNA with a high degree of specificity due to their large (e.g., >14 bp) cleavage sites. While the specificity of the homing endonucleases for their target sites allows for precise targeting of the induced DNA breaks, homing endonuclease cleavage sites are rare and the probability of finding a naturally occurring cleavage site in a targeted gene is low.
One class of artificial endonucleases is the zinc finger endonucleases. Zinc finger endonucleases combine a non-specific cleavage domain, typically that of FokI endonuclease, with zinc finger protein domains that are engineered to bind to specific DNA sequences. The modular structure of the zinc finger endonucleases makes them a versatile platform for delivering site-specific double-strand breaks to the genome. One limitation of the zinc finger endonucleases is that low specificity for a target site or the presence of multiple target sites in a genome can result in off-target cleavage events. As Fok1 endonuclease cleaves as a dimer, one strategy to prevent off-target cleavage events has been to design zinc finger domains that bind at adjacent 9 base pair sites.
Another class of artificial endonucleases is the engineered meganucleases. Engineered homing endonucleases are generated by modifying the specificity of existing homing endonucleases. In one approach, variations are introduced in the amino acid sequence of naturally occurring homing endonucleases and then the resultant engineered homing endonucleases are screened to select functional proteins which cleave a targeted binding site. In another approach, chimeric homing endonucleases are engineered by combining the recognition sites of two different homing endonucleases to create a new recognition site composed of a half-site of each homing endonuclease.
The mutagenicity of the double strand DNA breaks generated by both the naturally occurring and artificial endonucleases depend upon the precision of DNA repair. The double strand breaks caused by endonucleases are commonly repaired through non-homologous end joining (NHEJ), which is the major DNA double-strand break repair pathway for many organisms. NHEJ is referred to as “non-homologous” because the break ends are ligated directly without the need for a homologous template, in contrast to homologous recombination, which utilizes a homologous sequence to guide repair. Imprecise repair through this pathway can result in mutations at the break site, such as DNA base deletions and insertions as well as translocations and telomere fusion. When the mutations are made within the coding sequence of a gene, they can render the gene and its subsequent protein product non-functional, creating a targeted gene disruption or “knockout” of the gene.
Double strand DNA break repair through the NHEJ pathway is often not mutagenic. The majority of endonuclease-induced breaks repaired by the NHEJ pathway involve precise re-ligation, resulting in the restoration of the original DNA sequence. This is especially true of the types of DNA breaks created by the current endonuclease platforms available for engineering site-specificity, namely homing endonucleases (meganucleases) and zinc finger nucleases. Both of these types of enzymes leave compatible base pair overhangs that do not require processing prior to re-ligation by the NHEJ pathway. When the overhangs are compatible, NHEJ repairs the break with a high degree of accuracy. Thus, from a genome engineering standpoint, many of the cleavage events generated by the current site-specific endonuclease platforms are unproductive.
The need for additional solutions to these problems is manifest.